
George Profenza
1.7K posts

George Profenza retweeté

If you're a cracked infra engineer who:
- loves Go, ETL, OLAP DBs, k8s
- wants to change the future of video AI
- loves seeing their work translate to $$$
we're aggressively hiring right now @sievedata.
Dm me. 💜🚀
English
George Profenza retweeté

Another digital human / digital avatar test, this time using a LORA trained Wan2.1 model on 27 images.
(I've used ComfyUI for inference/tests. This is WIP: just 14 epochs trained)
#wan21 #machinelearning #diffusion #comfyui #avatar
GIF
English

The dataset was generated using VHAP ( github.com/ShenhanQian/VH… ). (The mesh is about 10K triangles)

English

First tests trying out Gaussian Avatars ! Congratulations to the teams at @TU_Muenchen and @Woven_Toyota for the amazing work !
project: shenhanqian.github.io/gaussian-avata…
paper: arxiv.org/abs/2312.02069
code: github.com/ShenhanQian/Ga…
#avatar #gaussiansplatting #machinelearning
English

@AmandaAskell I don’t know enough about law, however my intuition tells me, yes. Bob should instead of killing Pete, argue his wrong conviction and get paid for his 30 years. Depending on the system, Pete might have a tough time arguing he’s legally alive 😅 theguardian.com/world/2018/mar…
English
George Profenza retweeté

Excited to have a case study for our largest format project EVER finally onlinez 10m^2 of pixel madness😱📺✨made with WebGL and @nodes_io.
variable@variable_io
Infinite Dreams is real-time generative art and one of Europe’s largest LED screen installations, based at the flagship @hm store in Regent Street, London, created in collaboration with @hirschandmann. Read more on how we brought this 10x10m canvas to life. 1/n
English

@BryanChrisBrown @willeastcott If you inspect your .ply file with a binary/hex viewer you can double check the header section which is ASCII. Ideally the .ply will have the properties a trained 3DGS (x, y, z, f_dc0, f_dc1, f_dc2, opacity, scale_0, scale_1, scale_2, rot_0, rot_1, rot_2, rot_3). HTH
English

Announcing SuperSplat 📢
A new open-source tool for working with #GaussianSplatting scenes:
✂️Crop/delete splats
🔄Rotate splats
💾Resave
Links in 🧵
English

@JoshuaDavis @ProcessingOrg You can try rendering to SVG instead of canvas. I’ve used github.com/zenozeng/p5.js… in the past. HTH
English

The Omega Code / Finished up some final details for the top level layers. 100% all vector... Using @ProcessingOrg p5js, runs all in setup. Does anyone know if it's easy to export back to vector PDF from the browser... While using context radial gradients?


English

@drv_kmr has been working hard on his first Kickstarter campaign: a beautiful programmable light piece. Its controllable via #dmx (#streamingacn #artnet) as well as #mqtt Home Automation. It uses temp, CO2, humidity sensors driving Tamagotchi like interactions with simulated fish
Dhruv Kumar@drv_kmr
Livegrid is published on @Hacksterio with a heavy headline🙃. We need more exposure like this, please share Livegrid in your network and help us reach the right people. 25 days to go ! kickstarter.com/projects/liveg… #IoT #digitalart #environmentalawareness #homeassistant #opensource
English

@AndyTWoods A pleasure meeting you and the team! I do remember the creativity as3 required to get past limitations, but I also appreciate the freedom to apply the same energy to other creative coding toolkits like Processing, OpenFrameworks, etc
English

I'm fortunate to be giving a talk on the @opencvlibrary Webinar today to talk about Invisible: the Markerless Motion Capture solution by @disguise_one and move.ai.
It will not to technical,but hopefully fun
Feel free to join me in a few hours: 9AM PDT / 5PM GMT
OpenCV - Open Source Computer Vision Library@opencvlive
This Thursday on OpenCV Live we'll get a visit from OpenCV community member George Profenza, who will be showing us the new hotness in motion capture: It's called "Invisible"
English

@yaxu @RebeccaFiebrink Similarly Gesture Recognition Toolkit (GRT) might be worth a try (github.com/nickgillian/grt)
English

@yaxu Might a bit late to the party, but Dynamic Time Warping (DTW) comes to mind. I remember @RebeccaFiebrink's Wekinator handles this already (using OSC to get data in/out). I'd suspect you'd need to show it examples of each motion to 'count' separately first, then count repetitions
English
George Profenza retweeté

I'm doing a Reddit AMA with @mattleacock at 12 noon EDT // 5pm BST today (aka now!)
We'll be talking about the Daybreak crowdfunder, but ask ask us anything!
reddit.com/r/boardgames/c…
English

@TheKevinWorkman Thanks for the shout-out Kevin! (And thank you for contributing happycoding.io more and more I find myself sending people your How to Program/ How to debug articles) ❤️
English

side note I didn't do this on purpose but I love that the thumbnail that youtube chose was @orgicus's webpage. Check his stuff out here! sensori.al
English
George Profenza retweeté





